Storelectric Ltd

Project developers of grid-scale

Compressed Air Energy Storage (CAES)

The need for grid-scale energy storage

For more details on this subject, please enquire for our white papers.


The intermittency of renewables creates a need for additional power (demand surge) when the wind doesn’t blow and/or the sun doesn’t shine and when there is a need for less power (a demand trough) when it is sunny/windy – the imbalance between matching intermittent generation with variable demand is grid-scale. The advantages of grid-scale energy storage:

  • Both produces and stores electricity to balance the system and is therefore twice as versatile as gas plants;
  • Is not diminished by coincident weather patterns and/or demand in neighbouring countries with which there are interconnectors;
  • Can store at predictable peak production times allowing renewable energy to keep on producing and baseload electricity generation to continue unabated at maximum efficient levels, avoiding costly shutdowns or negative electricity prices; and
  • Provides the greatest level of energy security: can store electricity in country, generated in country from renewable resources.


As a result, millions of pounds are paid each year to stop wind farms producing electricity during trough demand – and not only wind farms, but mainly other generators, though these other generation curtailments are largely because of variable wind and solar generation. Millions more are paid to keep power stations (typically gas fired) operating at below 100% efficiency as a “hot standby” that can be fired up at a moment’s notice for peaks and surges. Millions more are paid (in discounted energy charges) to various energy users to permit the generators and grids to switch off parts of their consumption, to reduce the load during peaks and surges, accounting for about 2,000MW costing several thousand pounds per megawatt of capacity regardless of whether or not that capacity is used. They also keep about 500MW of diesel generators and 150MW of gas generators for standby service. Peak energy sells for up to 5 times the normal wholesale price of electricity.


Currently the system is backed up principally by using a combination of pumped hydro-electric storage and gas-fired power stations. This not only reduces dramatically the power that such power stations can sell, but also increases their cost and pollution in doing so, owing to frequent start-ups / ramp-ups and switch-offs / slow-downs. As a result they have become unprofitable, as a result of which all German gas-fired power stations were slated for closure until floor prices were introduced recently, and no new gas-fired power stations are being built in the UK without subsidy so the Capacity Market was developed to incentivise (or subsidise) them - so far, unsuccessfully.


Applying actual winds from January 2000 to the forecast wind farm generation capacity in 2030, and comparing this with the forecast mix of other generation technologies, shows the scale of the looming problem: while the country will need ~60GW of non-wind capacity for windless days such as 24-27 Jan, most of this is (expensively) unused most of the time. CCGT, for example, only has half of its capacity used four times in the month. However when the wind does blow, even much of the baseload nuclear capacity (which should never be stopped) has to be stopped 6 times.





The British government’s TINA (Technology Innovation Needs Analysis) identifies 27.4GW, 128GWh (i.e. average 5 hours’ duration) storage requirements by 2050.


Our own analysis suggests a need for 17-20GW in the UK. Across the EU the need is about 10 times this, and globally about 100 times to an estimated 7TW of capacity, or 3,500 plants of 500MW. And this is just to support intermittent generation, without considering supporting renewables for baseload or for helping supplant hydrocarbons in heating and transportation, which would again increase the market by a further 3-6 times for baseload, and ~10 times for transportation and heating.

National Grid’s Two Degrees scenario forecasts 73.3GW actual demand (plus 5-10% capacity margin required = 85GW demand) by 2040 to be satisfied by just 9.4GW of dispatchable (on-demand) and baseload generation of proven, likely technologies. The remaining 63GW is unreliable: intermittent renewables, interconnectors (neighbouring countries’ peaks mostly coincide with ours), CCS (all projects cancelled as too expensive) and nuclear (expensive). But even the demand level of 73.3GW appears to be a gross under-estimate.


For more details, see Matching the Solution to the Problem.

Alternatives to CAES

Once Britain (or any other country, for that matter) has sufficient energy storage we can power the ever increasing demands of the entire country from renewable sources, protecting the environment and reducing dependency on depleting fossil fuels, pollution and greenhouse gas emissions. Renewables and large scale storage can also contribute to replacing other fuels, such as:

  • Gas for heating and industrial processes, and
  • Liquid fuels for transportation.


Many grids consider gas-fired power stations to be the technology of choice for balancing renewables. These would operate most efficiently generating a constant baseload power, but which are used to deliver variable power. When we will have built sufficient energy storage capacity, these can be run constantly with most (if not all) required variation being drawn from the energy storage sites.


Apart from pumped hydroelectric storage, none of the alternatives is at grid scale (gigawatts and gigawatt-hours).

  • Hydroelectric pumped storage, in which two reservoirs at high and low level are linked. Water is pumped up during troughs to be released during peaks. About 300 such schemes have been installed world-wide, including Dinorwig (1,728 MW) and Ffestiniog (360MW) in Wales and Cruachan Dam and Foyers in Scotland; two further schemes are under construction in Scotland. However there are not thought to be many more suitable sites available, and those that are suitable are also subject to environmental objections for flooding scenic valleys. They are also very expensive; Dinorwig cost £425m to build in 1974 – equivalent to £3.75bn now. Battery storage has been proposed, but no technology has been demonstrated that offers grid-scale (multi-megawatt) storage capacity. It also suffers from efficiencies of 50-75% depending on the technology: in general, cost rises with efficiency.
  • Cryogenic air or liquid air energy storage is up to a maximum 70% efficiency, but with both high capex cost per MW and per MWh and scalability is limited to tens of megawatt-hours.
  • Flywheels have been proposed, but the only grid scale plants existing store 20MW for 15 minutes only and, despite state loan guarantees of $43m, the company developing it went bankrupt though it is operating again. It is not scalable practically to truly grid scale.
  • Hydrogen fuel cell energy storage is still under development, more for vehicles than for grid connected applications, with efficiencies of around 20-40%. There are also issues of scalability and of the process efficiency of electrolysing water to separate out the hydrogen is itself.
  • Various kinds of battery (flow or electrolytic) are being developed, limited to tens of megawatts, for minutes: most battery installations have a 30-minute duration at rated capacity, with some up to 2 hours. The energy they store is limited to megawatt-hour and costly to scale. There are also questions about the availability of scarce resources and the pollution created in extracting and transporting them, and in manufacturing the batteries.


© Storelectric 2013-16 All Rights Reserved